
import pandas as pd
pd.set_option('display.max_columns',None)
import numpy as np
import pandas_profiling
import matplotlib.pyplot as plt 
%matplotlib inline
import seaborn as sns
sns.set(style='whitegrid',color_codes=True)
import warnings
warnings.filterwarnings('ignore')
plt.rcParams['figure.figsize']=(10,7)


df=pd.read_csv("Telecome Churn.csv")


df.head()


df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 2666 entries, 0 to 2665
Data columns (total 20 columns):
 #   Column                  Non-Null Count  Dtype  
---  ------                  --------------  -----  
 0   State                   2666 non-null   object 
 1   Account length          2666 non-null   int64  
 2   Area code               2666 non-null   int64  
 3   International plan      2666 non-null   object 
 4   Voice mail plan         2666 non-null   object 
 5   Number vmail messages   2666 non-null   int64  
 6   Total day minutes       2666 non-null   float64
 7   Total day calls         2666 non-null   int64  
 8   Total day charge        2666 non-null   float64
 9   Total eve minutes       2666 non-null   float64
 10  Total eve calls         2666 non-null   int64  
 11  Total eve charge        2666 non-null   float64
 12  Total night minutes     2666 non-null   float64
 13  Total night calls       2666 non-null   int64  
 14  Total night charge      2666 non-null   float64
 15  Total intl minutes      2666 non-null   float64
 16  Total intl calls        2666 non-null   int64  
 17  Total intl charge       2666 non-null   float64
 18  Customer service calls  2666 non-null   int64  
 19  Churn                   2666 non-null   bool   
dtypes: bool(1), float64(8), int64(8), object(3)
memory usage: 398.5+ KB


df.describe()


df['State'].unique()

array(['KS', 'OH', 'NJ', 'OK', 'AL', 'MA', 'MO', 'WV', 'RI', 'IA', 'MT',
       'ID', 'VT', 'VA', 'TX', 'FL', 'CO', 'AZ', 'NE', 'WY', 'IL', 'NH',
       'LA', 'GA', 'AK', 'MD', 'AR', 'WI', 'OR', 'DE', 'IN', 'UT', 'CA',
       'SD', 'NC', 'WA', 'MN', 'NM', 'NV', 'DC', 'NY', 'KY', 'ME', 'MS',
       'MI', 'SC', 'TN', 'PA', 'HI', 'ND', 'CT'], dtype=object)


df['Area code'].unique()

array([415, 408, 510], dtype=int64)


df['International plan'].unique()

array(['No', 'Yes'], dtype=object)


df['Voice mail plan'].unique()

array(['Yes', 'No'], dtype=object)


df.isna().sum()

State                     0
Account length            0
Area code                 0
International plan        0
Voice mail plan           0
Number vmail messages     0
Total day minutes         0
Total day calls           0
Total day charge          0
Total eve minutes         0
Total eve calls           0
Total eve charge          0
Total night minutes       0
Total night calls         0
Total night charge        0
Total intl minutes        0
Total intl calls          0
Total intl charge         0
Customer service calls    0
Churn                     0
dtype: int64


df['Churn']=df['Churn'].replace({False:0,True:1})


df['Churn'].value_counts()

0    2278
1     388
Name: Churn, dtype: int64


df['Churn'].mean()

0.145536384096024


df.hist(figsize=(20,20),grid=False,color='pink',bins=15);


sns.countplot(data=df,x='Churn');


(df.groupby(['State'])['Churn'].mean()).nlargest(10).plot(kind='bar',color='green');


sns.countplot(data=df,x='Customer service calls',palette='Set3');


sns.countplot(data=df,x='Area code',palette='Accent_r')

<AxesSubplot:xlabel='Area code', ylabel='count'>


print(f'counts for International Plan \n {df["International plan"].value_counts()}\n\n')
print(f'counts for Voice Mail plan \n {df["Voice mail plan"].value_counts()}')

counts for International Plan 
 No     2396
Yes     270
Name: International plan, dtype: int64


counts for Voice Mail plan 
 No     1933
Yes     733
Name: Voice mail plan, dtype: int64


f,axes=plt.subplots(1,2,figsize=(15,8))
df["International plan"].value_counts().plot(kind='pie',autopct=lambda
                    p : '{:.0f}'.format(p * (df['International plan'].count()) / 100),wedgeprops=dict(width=.30),
                                             cmap='Pastel1',shadow=True,ax=axes[0])

df['Voice mail plan'].value_counts().plot(kind='pie',autopct=lambda
                                          p : '{:.0f}'.format(p * (df['Voice mail plan'].count()) / 100),shadow=True,
                                          wedgeprops=dict(width=.30),cmap='crest',ax=axes[1]);


print(f'counts for International Plan \n{df.groupby("International plan")["Churn"].value_counts()}\n\n')
print(f'counts for Voice Mail plan \n {df.groupby("Voice mail plan")["Churn"].value_counts()}\n')

counts for International Plan 
International plan  Churn
No                  0        2126
                    1         270
Yes                 0         152
                    1         118
Name: Churn, dtype: int64


counts for Voice Mail plan 
 Voice mail plan  Churn
No               0        1610
                 1         323
Yes              0         668
                 1          65
Name: Churn, dtype: int64


plt.figure(figsize=(15,8))
plt.subplot(1,2,1)
plt.title('International plan count respect to Churn')
df.groupby('International plan')['Churn'].value_counts().plot(kind='pie', autopct=lambda
                                     p : '{:.0f}'.format(p * (df['International plan'].count()) / 100), cmap='Set3_r',
                                       wedgeprops=dict(width=.3), shadow=True)

plt.legend(bbox_to_anchor=(1.3,1.25), ncol=2)

plt.subplot(1,2,2)
plt.title('Voice mail plan count respect to Churn')
df.groupby('Voice mail plan')['Churn'].value_counts().plot(kind='pie' ,autopct=lambda
                                        p : '{:.0f}'.format(p * (df['Voice mail plan'].count()) / 100), cmap='Set3_r',
                                              wedgeprops=dict(width=.3), shadow=True);


f,axes=plt.subplots(1,2,figsize=(15,8))

sns.barplot(data=df,x='International plan',y='Churn',estimator=np.mean,palette='viridis',ax=axes[0])

sns.barplot(data=df,x='Voice mail plan',y='Churn',estimator=np.mean,palette='summer',ax=axes[1]);

<AxesSubplot:xlabel='Voice mail plan', ylabel='Churn'>


sns.lmplot('Total day minutes', 'Total night minutes', data = df, hue='Churn', fit_reg=False,palette='PuRd',markers=['o','x'],size=8);


sns.lmplot('Total day calls', 'Total night calls', data = df, hue='Churn', fit_reg=False,palette='viridis',markers=['o','x'],size=8);


sns.pairplot(df,palette='Set2',hue='Churn')

<seaborn.axisgrid.PairGrid at 0x28c20c8f490>


plt.figure(figsize=(15,15))
sns.heatmap(df.corr(),annot=True,cmap='summer',linewidths=2,linecolor='orange',annot_kws={'size':8,'color':'black'},fmt='.4f',square=True)

<AxesSubplot:>


sns.boxplot(data=df,x='Churn',y='Total intl minutes',palette='PuRd')

<AxesSubplot:xlabel='Churn', ylabel='Total intl minutes'>


sns.boxplot(data=df,x='Area code',y='Total eve charge',palette='summer')

<AxesSubplot:xlabel='Area code', ylabel='Total eve charge'>


# df.groupby(['Churn'])['Total day charge'].mean().plot(kind='bar')
sns.barplot(data=df,y='Total day charge',x='Churn',palette='rainbow');


    sns.barplot(data=df,x='Area code',y='Customer service calls',estimator=sum,hue='Churn',palette='Set2');


sns.barplot(data=df,y='Total night minutes',x='Churn',palette='viridis');


df.head()


from sklearn.preprocessing import LabelEncoder,OneHotEncoder


df=pd.get_dummies(df) #OneHotEncoder


df.head()


from sklearn.preprocessing import RobustScaler


roscaler=RobustScaler()


df[['Account length','Area code','Number vmail messages','Total day minutes','Total day calls'
    ,'Total day charge','Total eve minutes','Total eve calls','Total eve charge','Total night minutes'
    ,'Total night calls','Total night charge','Total intl minutes','Total intl calls','Total intl charge'
    ,'Customer service calls']]=roscaler.fit_transform(df[['Account length','Area code','Number vmail messages','Total day minutes','Total day calls'
                            ,'Total day charge','Total eve minutes','Total eve calls','Total eve charge'
                            ,'Total night minutes','Total night calls','Total night charge','Total intl minutes'
                            ,'Total intl calls','Total intl charge','Customer service calls']])


features=df.drop(['Churn'],axis=1)


target=df['Churn']


from sklearn.model_selection import train_test_split


features_train,features_test,target_train,target_test=train_test_split(features,target,test_size=0.2,random_state=1)


from sklearn.linear_model import LogisticRegression,SGDClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier,AdaBoostClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.svm import SVC
from sklearn.naive_bayes import GaussianNB
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis, QuadraticDiscriminantAnalysis
from xgboost import XGBClassifier

from sklearn.metrics import confusion_matrix,classification_report,precision_score,recall_score


models={'LogisticRegression':LogisticRegression(),
        'SGDC':SGDClassifier(),
        'DecisionTree':DecisionTreeClassifier(),
        'RandomForest':RandomForestClassifier(),
        'GradientBoost':GradientBoostingClassifier(),
        'KNN':KNeighborsClassifier(),
        'SVM':SVC(),
        'AdaBoost':AdaBoostClassifier(),
        'Naive Bayes':GaussianNB(),
        'LinearDiscriminant':LinearDiscriminantAnalysis(),
        'QuadraticDiscriminant':QuadraticDiscriminantAnalysis(),
        'XGBoost':XGBClassifier()
       }


for name, model in models.items():
    model1=model.fit(features_train,target_train)
    y_pred = model1.predict(features_test)
    con_mat=pd.DataFrame(confusion_matrix(target_test,y_pred),index=['Actual Negetive','Actual Positive'],columns=['Predicted Negetive','Predicted Positive'])
    print(f"{name} confuion matrix:--------> \n\n\n {con_mat}\n\n\n") 
    print(f'{name} accuracy score:----> {model1.score(features_test,target_test)}\n')
    print(f'{name} precision score:----> {precision_score(target_test,y_pred)}\n')
    print(f'{name} recall score:----> {recall_score(target_test,y_pred)}\n')
    print(f'{name} classification report:-------> \n\n {classification_report(target_test,y_pred)}\n\n\n')

LogisticRegression confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 440                  19
Actual Positive                  58                  17


LogisticRegression accuracy score:----> 0.8558052434456929

LogisticRegression precision score:----> 0.4722222222222222

LogisticRegression recall score:----> 0.22666666666666666

LogisticRegression classification report:-------> 

               precision    recall  f1-score   support

           0       0.88      0.96      0.92       459
           1       0.47      0.23      0.31        75

    accuracy                           0.86       534
   macro avg       0.68      0.59      0.61       534
weighted avg       0.83      0.86      0.83       534


SGDC confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 438                  21
Actual Positive                  56                  19


SGDC accuracy score:----> 0.8558052434456929

SGDC precision score:----> 0.475

SGDC recall score:----> 0.25333333333333335

SGDC classification report:-------> 

               precision    recall  f1-score   support

           0       0.89      0.95      0.92       459
           1       0.47      0.25      0.33        75

    accuracy                           0.86       534
   macro avg       0.68      0.60      0.62       534
weighted avg       0.83      0.86      0.84       534


DecisionTree confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 433                  26
Actual Positive                  27                  48


DecisionTree accuracy score:----> 0.900749063670412

DecisionTree precision score:----> 0.6486486486486487

DecisionTree recall score:----> 0.64

DecisionTree classification report:-------> 

               precision    recall  f1-score   support

           0       0.94      0.94      0.94       459
           1       0.65      0.64      0.64        75

    accuracy                           0.90       534
   macro avg       0.79      0.79      0.79       534
weighted avg       0.90      0.90      0.90       534


RandomForest confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 459                   0
Actual Positive                  29                  46


RandomForest accuracy score:----> 0.9456928838951311

RandomForest precision score:----> 1.0

RandomForest recall score:----> 0.6133333333333333

RandomForest classification report:-------> 

               precision    recall  f1-score   support

           0       0.94      1.00      0.97       459
           1       1.00      0.61      0.76        75

    accuracy                           0.95       534
   macro avg       0.97      0.81      0.86       534
weighted avg       0.95      0.95      0.94       534


GradientBoost confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 451                   8
Actual Positive                  21                  54


GradientBoost accuracy score:----> 0.9456928838951311

GradientBoost precision score:----> 0.8709677419354839

GradientBoost recall score:----> 0.72

GradientBoost classification report:-------> 

               precision    recall  f1-score   support

           0       0.96      0.98      0.97       459
           1       0.87      0.72      0.79        75

    accuracy                           0.95       534
   macro avg       0.91      0.85      0.88       534
weighted avg       0.94      0.95      0.94       534


KNN confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 454                   5
Actual Positive                  54                  21


KNN accuracy score:----> 0.8895131086142322

KNN precision score:----> 0.8076923076923077

KNN recall score:----> 0.28

KNN classification report:-------> 

               precision    recall  f1-score   support

           0       0.89      0.99      0.94       459
           1       0.81      0.28      0.42        75

    accuracy                           0.89       534
   macro avg       0.85      0.63      0.68       534
weighted avg       0.88      0.89      0.87       534


SVM confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 448                  11
Actual Positive                  39                  36


SVM accuracy score:----> 0.9063670411985019

SVM precision score:----> 0.7659574468085106

SVM recall score:----> 0.48

SVM classification report:-------> 

               precision    recall  f1-score   support

           0       0.92      0.98      0.95       459
           1       0.77      0.48      0.59        75

    accuracy                           0.91       534
   macro avg       0.84      0.73      0.77       534
weighted avg       0.90      0.91      0.90       534


AdaBoost confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 433                  26
Actual Positive                  48                  27


AdaBoost accuracy score:----> 0.8614232209737828

AdaBoost precision score:----> 0.5094339622641509

AdaBoost recall score:----> 0.36

AdaBoost classification report:-------> 

               precision    recall  f1-score   support

           0       0.90      0.94      0.92       459
           1       0.51      0.36      0.42        75

    accuracy                           0.86       534
   macro avg       0.70      0.65      0.67       534
weighted avg       0.85      0.86      0.85       534


Naive Bayes confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 282                 177
Actual Positive                  31                  44


Naive Bayes accuracy score:----> 0.6104868913857678

Naive Bayes precision score:----> 0.19909502262443438

Naive Bayes recall score:----> 0.5866666666666667

Naive Bayes classification report:-------> 

               precision    recall  f1-score   support

           0       0.90      0.61      0.73       459
           1       0.20      0.59      0.30        75

    accuracy                           0.61       534
   macro avg       0.55      0.60      0.51       534
weighted avg       0.80      0.61      0.67       534


LinearDiscriminant confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 431                  28
Actual Positive                  53                  22


LinearDiscriminant accuracy score:----> 0.848314606741573

LinearDiscriminant precision score:----> 0.44

LinearDiscriminant recall score:----> 0.29333333333333333

LinearDiscriminant classification report:-------> 

               precision    recall  f1-score   support

           0       0.89      0.94      0.91       459
           1       0.44      0.29      0.35        75

    accuracy                           0.85       534
   macro avg       0.67      0.62      0.63       534
weighted avg       0.83      0.85      0.84       534


QuadraticDiscriminant confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 317                 142
Actual Positive                  34                  41


QuadraticDiscriminant accuracy score:----> 0.6704119850187266

QuadraticDiscriminant precision score:----> 0.22404371584699453

QuadraticDiscriminant recall score:----> 0.5466666666666666

QuadraticDiscriminant classification report:-------> 

               precision    recall  f1-score   support

           0       0.90      0.69      0.78       459
           1       0.22      0.55      0.32        75

    accuracy                           0.67       534
   macro avg       0.56      0.62      0.55       534
weighted avg       0.81      0.67      0.72       534


[14:39:36] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.5.1/src/learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'binary:logistic' was changed from 'error' to 'logloss'. Explicitly set eval_metric if you'd like to restore the old behavior.
XGBoost confuion matrix:--------> 


                  Predicted Negetive  Predicted Positive
Actual Negetive                 457                   2
Actual Positive                  20                  55


XGBoost accuracy score:----> 0.9588014981273408

XGBoost precision score:----> 0.9649122807017544

XGBoost recall score:----> 0.7333333333333333

XGBoost classification report:-------> 

               precision    recall  f1-score   support

           0       0.96      1.00      0.98       459
           1       0.96      0.73      0.83        75

    accuracy                           0.96       534
   macro avg       0.96      0.86      0.90       534
weighted avg       0.96      0.96      0.96       534


XGBoost=XGBClassifier()


XGBoost.fit(features_train,target_train)

[14:42:46] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.5.1/src/learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'binary:logistic' was changed from 'error' to 'logloss'. Explicitly set eval_metric if you'd like to restore the old behavior.

XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,
              colsample_bynode=1, colsample_bytree=1, enable_categorical=False,
              gamma=0, gpu_id=-1, importance_type=None,
              interaction_constraints='', learning_rate=0.300000012,
              max_delta_step=0, max_depth=6, min_child_weight=1, missing=nan,
              monotone_constraints='()', n_estimators=100, n_jobs=4,
              num_parallel_tree=1, predictor='auto', random_state=0,
              reg_alpha=0, reg_lambda=1, scale_pos_weight=1, subsample=1,
              tree_method='exact', validate_parameters=1, verbosity=None)


XGBoost.score(features_test,target_test)

0.9588014981273408


import pickle
Telecom_churn_model=open('Telecom_churn.sav','wb')
pickle.dump(XGBoost,Telecom_churn_model)


from sklearn.cluster import DBSCAN,AgglomerativeClustering,KMeans
from sklearn.metrics import silhouette_score


model=KMeans(2,random_state=2)


predicted=model.fit_predict(features)


predicted=pd.DataFrame(predicted,columns=['predicted'])


new_clustered_df=pd.concat([features,target,predicted],axis=1)


new_clustered_df[['Churn','predicted']]


silhouette_score(features,predicted)

0.09471835154480508


model=AgglomerativeClustering()


predicted=model.fit_predict(features)


silhouette_score(features,predicted)

0.09951184912546202

	Account length	Area code	Number vmail messages	Total day minutes	Total day calls	Total day charge	Total eve minutes	Total eve calls	Total eve charge	Total night minutes	Total night calls	Total night charge	Total intl minutes	Total intl calls	Total intl charge	Customer service calls
count	2666.000000	2666.000000	2666.000000	2666.00000	2666.000000	2666.000000	2666.000000	2666.000000	2666.000000	2666.000000	2666.000000	2666.000000	2666.000000	2666.000000	2666.000000	2666.000000
mean	100.620405	437.438860	8.021755	179.48162	100.310203	30.512404	200.386159	100.023631	17.033072	201.168942	100.106152	9.052689	10.237022	4.467367	2.764490	1.562641
std	39.563974	42.521018	13.612277	54.21035	19.988162	9.215733	50.951515	20.161445	4.330864	50.780323	19.418459	2.285120	2.788349	2.456195	0.752812	1.311236
min	1.000000	408.000000	0.000000	0.00000	0.000000	0.000000	0.000000	0.000000	0.000000	43.700000	33.000000	1.970000	0.000000	0.000000	0.000000	0.000000
25%	73.000000	408.000000	0.000000	143.40000	87.000000	24.380000	165.300000	87.000000	14.050000	166.925000	87.000000	7.512500	8.500000	3.000000	2.300000	1.000000
50%	100.000000	415.000000	0.000000	179.95000	101.000000	30.590000	200.900000	100.000000	17.080000	201.150000	100.000000	9.050000	10.200000	4.000000	2.750000	1.000000
75%	127.000000	510.000000	19.000000	215.90000	114.000000	36.700000	235.100000	114.000000	19.980000	236.475000	113.000000	10.640000	12.100000	6.000000	3.270000	2.000000
max	243.000000	510.000000	50.000000	350.80000	160.000000	59.640000	363.700000	170.000000	30.910000	395.000000	166.000000	17.770000	20.000000	20.000000	5.400000	9.000000

	Account length	Area code	Number vmail messages	Total day minutes	Total day calls	Total day charge	Total eve minutes	Total eve calls	Total eve charge	Total night minutes	Total night calls	Total night charge	Total intl minutes	Total intl calls	Total intl charge	Customer service calls	State_KS	State_NJ	State_OH	State_OK	International plan_No	International plan_Yes	Voice mail plan_No	Voice mail plan_Yes
0	128	415	25	265.1	110	45.07	197.4	99	16.78	244.7	91	11.01	10.0	3	2.70	1	1	0	0	0	1	0	0	1
1	107	415	26	161.6	123	27.47	195.5	103	16.62	254.4	103	11.45	13.7	3	3.70	1	0	0	1	0	1	0	0	1
2	137	415	0	243.4	114	41.38	121.2	110	10.30	162.6	104	7.32	12.2	5	3.29	0	0	1	0	0	1	0	1	0
3	84	408	0	299.4	71	50.90	61.9	88	5.26	196.9	89	8.86	6.6	7	1.78	2	0	0	1	0	0	1	1	0
4	75	415	0	166.7	113	28.34	148.3	122	12.61	186.9	121	8.41	10.1	3	2.73	3	0	0	0	1	0	1	1	0

	Churn	predicted
0	0	0
1	0	0
2	0	0
3	0	1
4	0	1
...	...	...
2661	0	1
2662	0	1
2663	0	1
2664	0	1
2665	0	0

Telecom Churn Analysis¶

Importing Libraries¶

Reading Dataset¶

Description of Columns¶

Getting Information About Dataset¶

Basic Statistics of Data¶

Checking for Unique Values for States , Area Code , International plan , Voice mail plan¶

Cleaning¶

Checking for Null Values¶

There is No Null Values in Dataset¶

So, We can go for EDA¶

For Target Replacing True and False with 0 & 1 Accordingly¶

Probability of Custumer Left The Services is 14%¶

Plotting Historams for All the Columns¶

Here from Histogram we can say that in Columms of Account length , Total day minutes , Total day calls , Total day charge , Total eve minutes ,Total eve calls, Total eve charg , Total night minutes , Total night calls , Total night charge , Total intl minutes & Total intl charge

Most Data Lie Near Mean & Median

Countplot of Churn¶

Less than 500 Customers Left the Services¶

Top 10 States Who Left the Sevices¶

Graph Shows Probability of State who Left The Services¶

Countplot for Customer Service Calls¶

Countplot for Customer Area Code¶

Counts of Customes Having International Plans & Voice Mail Plan¶

Count of Customer Who Purchased International Plajn / Voice mail Plan with Respect to Churn¶

1.Probability of Leaving Services Who have International Plan or Not¶

2.Probability of Leaving Services Who have Voice Mail Plan or Not¶

Scatterplot of Total day minutes & Total night minutes with respect to Churn¶

Scatterplot of Total day Calls & Total night Calls with respect to Churn¶

Pairplot of Data¶

Getting Correlation By Heatmap¶

Boxplot of Total International Minutes¶

Boxplot of Total Evening Minutes¶

Average Total Day Charges with Respect To Churn¶

The Charges are Higher for Customers who Left Services¶

There is no such Difference in Usage of Minutes in night¶

We can go for Machine learning Model¶

Here some columns are Categorical So, Label encoding is Required¶

Scaling Data¶

Defining Objects For Features And Target¶

Splitting Data for Training And Testing¶

Trying All Models on Data For Getting Best Accuracy¶

Best Accuracuy In Classification Algorithms is 96% in XgBoost and f1 Score is 0.83¶

Dumping Model¶

Algorithms for Clustering¶

Kmeans Clustering¶

Checking silhouette_score¶

Hirearchical Clustering¶